# OpenAI, Google and Microsoft back a shared AI opt-out signal

> A proposed website setting would let publishers signal AI training preferences to multiple companies through robots.txt, rather than configuring each one separately.

Oossa · 2026-10-30 · https://oossa.com/en/openai-google-and-microsoft-back-a-shared-ai-opt-out-signal

OpenAI, Google, Microsoft and Anthropic say they will support a shared signal that publishers can use to tell AI crawlers not to access their sites. The proposed “policy token” would sit in a website’s robots.txt file, a text file that already tells web crawlers which parts of a site they may visit. The aim is to let publishers set preferences once for several AI companies instead of dealing with each company separately.

The signal is being proposed within the Internet Engineering Task Force (IETF), a group that develops internet standards. It comes from Kevin Kelley, a technical lead for standards at Anthropic. Anthropic has already adopted the signal. Google and Microsoft plan to support it in 2027, while OpenAI says it will implement it in the coming months. The companies say their AI crawlers will restrict access when they encounter the token.

## What can publishers choose?

Microsoft says it will offer more detailed controls: publishers will be able to opt out of AI training, AI grounding, or both. Grounding means directing an AI model to a verified information source to help reduce incorrect answers. Microsoft says a site’s preferences will not remove it from Bing search results, separating AI uses of a site’s content from search indexing.

The announcement arrives as publishers face continued scraping by AI crawlers. Current arrangements require websites to signal AI training preferences separately to each company. The shared token could reduce that repetitive setup for publishers, but its usefulness depends on AI services recognizing and following the signal. The announcement does not spell out the token’s detailed format or explain how publishers will verify that every service has applied their preferences.

## Other approaches to scraping

The policy token joins other efforts to give publishers more control over how crawlers use their material. The Really Simple Licensing standard, or RSL, gives publishers a way to charge crawlers that visit their sites. Cloudflare, meanwhile, blocks AI crawlers by default. These approaches differ: the proposed token communicates a preference across participating companies, while RSL focuses on licensing and Cloudflare on blocking crawler access.

For now, publishers will have to account for different adoption schedules, and Microsoft’s promised controls go beyond a simple training opt-out. The shared approach is intended to replace separate signals to each company, but the companies’ commitments are not all in place yet.

## The facts

- The proposed policy token would be placed in a site’s robots.txt file.
- Anthropic says it has already adopted the signal; Google and Microsoft plan to support it in 2027.
- OpenAI says it will implement the signal in the coming months.
- Microsoft says publishers can separately opt out of AI training or AI grounding, while remaining in Bing search results.

## Why it matters

For publishers, a shared signal could mean setting AI-use preferences in one place instead of repeating them for each participating company. Whether that will cover the services they care about depends on adoption; Google and Microsoft’s support is still planned, not already available.

## Sources & references

1. [The biggest names in AI are backing a new way to opt out of scraping](https://www.theverge.com/tech/1023763/openai-google-microsoft-ai-scraping-standard-opt-out) – The Verge, 2026-10-30
2. [Google, Anthropic and others to support new AI crawling opt-out standard](https://siliconangle.com/2026/10/30/google-anthropic-and-others-to-support-new-ai-crawling-opt-out-standard/) – SiliconANGLE, 2026-10-31

Last updated: 2026-10-31